Textual vs. Iconic Programming Languages: Differences in Student Performance and Attitudes
Jing Lin, Mei-Ching Yang, Sheue-Yun Chen · 2007
Fifty-two sixth graders from a local elementary school participated in this study. Half of them learned to program in MSWLogo and the other half learned Drape. Both MSWLogo and Drape are Logo-style languages except that MSWLogo is a textual language while Drape is iconic. An analysis of students' test scores revealed that the MSWLogo group performed significantly better than the Drape group in the first three written quizzes. However, there was no significant difference found between the two groups in the fourth quiz, the final written test, or the hands-on programming exam administered at the end of the experiment. It seemed that, contrary to common perception, Drape (an iconic language) had a steeper learning curve than MSWLogo (a textual language). However, when students became more and more familiar with the language, the difference in student performance narrowed. With respect to usability of either language, students of the MSWLogo group made syntactic errors frequently while students of the Drape group were often confused by icons that were not intuitively clear or icons that looked similar. Again, the iconic language did not offer advantages in usability to the elementary school students as one would naturally expect. Regarding students' attitudes toward programming, no significant difference was found between the two groups in how much they enjoyed programming or if they would like to learn to program again.